Amino-enzyme intermediates in pepsin-catalyzed reactions

Amino-enzyme intermediates in pepsin-catalyzed reactions. Marc S. Silver, and Mai Stoddard. Biochemistry , 1972, 11 (2), pp 191–200. DOI: 10.1021/ ...
0 downloads 0 Views 1MB Size
PEPSIN MECHANISMS

Amin-Enzyme

Intermediates in Pepsin-Catalyzed Reactions”

Marc S. Silver? and Mai Stoddard

ABSTRACT :

The pepsin-catalyzed hydrolysis of Ac-Phe-Tyr at pH 4.5-4.7 produces an intermediate which reacts with radioactive Ac-Phe (Ac-Phe) to generate Ac-Phe-Tyr. The intermediate undergoes hydrolysis to Tyr about three times more readily than it reacts with 2.4 x lo-* M Ac-Phe. AcPhe-Phe behaves similarly to Ac-Phe-Tyr, but neither AcPhe-Tyr-NHz nor Ac-Phe-Phe-OEt yields an intermediate which Ac-Phe can trap. An amino-enzyme mechanism which postulates that pepsin-catalyzed hydrolyses follow the sequence [enzyme-substrate complex + amino-enzyme Ac-Phe + final products] appears incapable of explaining these observations in a simple manner. It also cannot account for the quantitative difference in the rate of incorporation of free Ac-Phe into unreacted Ac-Phe-Tyr as measured

+

A

11 efforts to isolate amino-enzymes from the pepsincatalyzed hydrolysis of peptides have failed (Kitson and Knowles, 1970). The premise that peptic hydrolyses normally produce such intermediates (eq 1) therefore rests upon two kinds of indirect evidence. Transpeptidation experiments, pioneered by Neumann et a f . (1959) and Fruton et af. (1961), suggest that an acceptor X’COO- can react with the aminoenzyme from XCONHY to produce X’CONHY while kinetic experiments establish that eq 1 accounts for the inhibitor E

+ ‘NH3Y

11 11

XCONHY

slow + E e E,XCONHY e +[E-NHY] + xcoo-

X’CONHY

+Ee

11

[E-NHY]

+

X’COO-

(1)

action of hydrolysis products and their analogs (Kitson and Knowles, 1971a,b). Numerous other experimental results are compatible with the amino-enzyme hypothesis but have revealed nothing about the nature of these intermediates (Fruton, 1970).

* From the Department of Chemistry, Amherst College, Amherst, Massachusetts 01002. Receiued August 9, 1971. Supported by Grant AM-08005 of the U. S.Public Health Service. This research, at its inception, benefited from the advice of members of the Department of Biophysics at the Weizmann Institute, particularly Dr. H. Neumann, and from an NSF Senior Postdoctoral Fellowship to M. S.S. t T o whom to address correspondence. 1 Ginodman et al. (1971) have recently claimed isolation of pepsin with L-tyrosine ethyl ester covalently attached.

directly with Ac-Phe and indirectly by the Ac-Phe-Tyrinduced exchange of ‘*OHZ with Ac-PheCOOH. Modifications of the simple amino-enzyme mechanism o r a mechanism which assumes that hydrolysis proceeds ziia two covalent enzyme-substrate intermediates lead to no really satisfactory rationalization of the experimental data. It seems probable that the transpeptidation reaction is only characteristic of the hydrolysis at high pH of synthetic substrates with a C-terminal carboxyl group near the point of bond cleavage. If so, mechanistic generalizations based upon the transpeptidation reaction are unjustified until the occurrence of amino-enzyme intermediates in the hydrolysis of substrates like Ac-Phe-Tyr-NHZ (a better model than Ac-PheTyr for a polypeptide) is established.

A fundamental property of each amino-enzyme should be its partitioning ratio, PR, the ratio of its rate of transpeptidation to its rate of hydrolysis under specified conditions. We decided to determine some PR’s, in the expectation that this knowledge of them should ultimately lead to an understanding of the properties of the amino-enzyme intermediates. The unexpected results obtained have raised serious doubts about the generality of the simple amino-enzyme concept, as exemplified by eq 1. The following five considerations guided our experimental approach and led us to use Ac-Phe-Tyr and Ac-Phe-Tyr-NHz as primary sources of amino-enzymes and Ac-Phe (radioactive Ac-Phe) as acceptor : (1) the available detailed kinetic data on the hydrolysis of each substrate (Denburg et al., 1968) would facilitate analysis of the transpeptidation results ; (2) the behavior of the amino-enzyme “E-NH-TyrCONHl’ from Ac-Phe-Tyr-NHZ might be more readily interpreted than that of “E-NH-TyrCOOH” from Ac-Phe-Tyr, since the former lacks an ionizable substrate carboxyl group; (3) the choice of Ac-Phe as acceptor must further ease analysis of the experimental observations, for transpeptidation generates only radioactive substrate, not a new substance (cf. Fruton et al., 1961); (4) the high solubility of Ac-Phe permits its use at high concentration, which should enhance the extent of incorporation of radioactivity into unreacted substrate and suppress undesirable transpeptidations in which the substrate also acts as acceptor (Neumann et af.,1959); and ( 5 ) we could compare our determination of the extent of exchange between Ac-Phe-Tyr and Ac-Phe to that of Kozlov et af. (1965) who measured the degree to which Ac-Phe-Tyr increased the rate of exchange of Ac-PheCOOH with l80Hz (each transformation XCONHY --t X’CONHY exchanges one oxygen of Ac-PheCOOH with solvent). %Abbreviationsare listed in Biochemistry 5, 2485 (1966), and all amino acids possess the L configuration unless otherwise specified. A functional group is sometimes added to the abbreviation for an amino acid to clarify a point (e.g., Ac-PheCOOH, +NHq-Tyr). Italicized Phe (Phe) designates 3H-labeled Phe.

B I O C H E M I S T R VYO, L . 1 1 ,

NO.

2, 1 9 7 2

191

SILVER A N D

Experimental Section Thin-Layer Chromatography (Tlc). Tlc analysis established the purity of amino acid derivatives and the qualitative or semiquantitative composition of reaction mixtures from enzymatic hydrolyses. All tlc's utilized silica gel G on glass plates with 4 : 1 :1 (Y/v) 1-butanol-acetic acid-water as the developing solvent. Typical RF values were: pepsin, 0.0; Tyr, 0.40; Tyr-Tyr, 0.59; Ac-Phe-Tyr, 0.73; Tyr-NHn,0.75. We employed the following visualization reagents of Stahl (1962) : Pauly (37) or Folin-Ciocalteau (122) for phenols; chlorine-tolidine (32) for peptide bonds; ninhydrin (108) for free amino acids. High- Voltage Paper Electrophoresis (Hue). A Savant Instruments Inc. FP30A instrument provided the fundamental tool for perf01 ming the separations necessary to this investigation. Samples of 10 p1 were spotted 3.8 cm apart on a piece of Whatman No. 3MM paper 27 cm wide (six samples per sheet). Electrophoresis proceeded for 2 hr at 4900 V (-150 Vjcm) and employed pH 5.3 buffer (20 ml of pyridine plus 8 ml of acetic acid diluted to 3 1). The paper was dried for 2 hr at 100" immediately after completion of the hve run and later cut into six 3.4-cm wide strips, each containing one sample. Strips were either sprayed (as for tlc) or cut at appropriate places into 1.7 X 1.7 cm squares (two squares/l.7 cm running length) for radioactivity determination. Radioactive spots on some strips destined for cutting were first located with a Tracerlab 4 a scanner. Counting Technique. Each scintillation vial was counted for 10 min or longer on a Packard 3314 Tri-Carb scintillation spectrometer and contained 11 ml of Bray's solution (Bray, 1960) plus either 10 pl of sample solution or a 1.7 X 1.7 cm square of unsprayed electrophoresis sheet ("paper sample") lying flat on the bottom of the vial. Controls revealed that allowing the paper samples to soak in the Bray solution for 12 hr prior to counting gave a reproducible, constant counting rate. The dual-channels ratio technique (Bush, 1963) established a counting efficiency of 15 % for both liquid and paper samples. Sprayed electrophor eograms showed much lower efficiencies. The following procedure guaranteed accurate determination of the radioactivity of paper samples. Two 10-pl samples of each solution applied to a given hve sheet were counted directly in Bray, while the total radioactivity of one strip (or two) of the electrophoreogram was measured by cutting all of it into squares. The total counts per minute for the paper samples therefore determined the counting efficiency for paper relative to liquid sample ("relative efficiency"). Calculations assumed all strips of an electrophoreogram possessed identical relative efficiency. Directly measured relative efficiencies for two strips of the same electrophoreogram a ways agreed to 3%. Relative efficiencies for a set of 36 strips from -30 different sheets averaged 91 =t 3 x with a range of 82-99z. Materials Purchased. Twice-crystallized pepsin was Worthington Biochemical Corp., lot PM 709. The following chemicals, purchased from Cyclo Chemical Corp. (CCC), Sigma Chemical Corp. (SCC), Fox Chemical Corp. (FCC), Calbiochem (CBC), or Aldrich Chemical Co. (ACC), had physical properties that agreed with literature values and were used as received : Ac-Phe-Tyr (CCC, mp 216-21 8 "), Tyr-Tyr .2H20 (CCC), Ac-Phe (CCC), Tyr-NHn.HC1 (SCC), D-Tyr-NHs (FCC), Tyr-Tyr-Tyr . 2He0 (FCC), Tyr (CBC), and Tyr-OMe (ACC). Ac-Phe-Tyr-NHz (mp 249-252") and Ac-Phe-TyrOMe (mp 114-116"; Clement et al. (1968) report mp 125-126 and 136-1 37"), obtained from Cyclo, were recrystallized

192

BIOCHEMISTRY,

VOL.

11,

NO.

2, 1 9 7 2

STODDARD

prior to use. Phe-OMe.HC1, a gift of Dr. M. S. Matta, had mp 160-162" (lit. mp 161-163" for the D isomer (Cohen and Schultz, 1968)). Ac-Phe-Phe-OEt was the gift of Dr. J. R. Knowles (Cornish-Bowden and Knowles, 1969). Materials Synthesized. Reaction of Ac-DL-Phe (mp 15& 153", lit. mp 152.5-153" (Overby and Ingersoll, 1951)) with Phe (Knowles et al., 1969) and fractional recrystallization of the resultant Ac-DL-Phe-L-Phe, mp 140-175" provided pure Ac-Phe-Phe (mp 245-250" (from methanol), [a]: +15.0" (C 0.6, pyridine)) and Ac-D-Phe-L-Phe (mp 178-194" (from ethyl acetate)). Knowles et al. report mp 245-249" and [a]:: +14.3" for the former and mp 193-195" for AC-LPhe-D-Phe. Reaction of Z-D-Phe, D-TYC-NH?, and isobutyl chloroformate (exp 4 of Anderson et al., 1967) yielded Z-D-Phe-D-TyrNH2, mp 190-195". Catalytic removal of the Z group, acetylation of the uncharacterized D-Phe-D-Tyr-NHs, and subsequent alkali treatment (Hofmann et al., 1960) gave Ac-D-Phe-D-Tyr-NHs, mp 249-252 ". The product behaved identically with its enantiomer on tlc and, when treated with pepsin, showed no detectable hydrolysis (ninhydrin reagent) after 19.5 hr under conditions where Ac-Phe-Tyr-NH2 showed 80% reaction in 2 hr and Ac-Phe-Tyr, 61 in 7 hr. Radiochemicals. The Radiochemical Centre, Amersham, England, reduced 5 mg of a-acetaminocinnamic acid, mp 190-192" (Herbst and Shemin, 1943), with tritium gas over a 10 % palladium-on-charcoal catalyst. They added 400 mg of unlabeled Ac-DL-Phe and supplied the resultant 600-mCi sample as a solution in 10 ml of ethanol, to be referred to as "stock Ac-DL-Phe." Some stock Ac-DL-Phe was purified for kinetic and synthetic use. A mixture of 8.0 g of unlabeled Ac-DL-Phe and 2 ml of stock Ac-DL-Phe gave 7.0 g of product upon recrystallization from hot water. Further recrystallization from acetone and then again from water gave 5.4 g of Ac-DL-Phe, mp 151-153". This sample was recrystallized from acetone approxirnately every 6 months and typically had mp 152.5-154", specific activity 3 X lo6 cpm/mg. Reaction of Ac-DL-Phe with Phe, as described above for the nonradioactive substance, provided Ac-Phe-Phe (mp 246-253") and Ac-D-Phe-L-Phe (mp 193194.5"). Enzymatic synthesis of Ac-Phe-Tyr and Ac-Phe-Tyr-OMe employed Ac-Phe as starting material. This was obtained from 3.5 g of Ac-Phe plus 0.5 ml of stock Ac-DL-Phe and had mp 166-167" and a constant specific activity of -7 X l o 5 cpm/mg after several recrystallizations from water and acetone. Incubation of 650 mg of Ac-Phe and 3.1 g of Tyr-OMe with pepsin at pH 4 (Kozlov et al., 1966) yielded crude Ac-Phe-Tyr-OMe which upon recrystallization from ethyl acetate-hexane gave a first crop of 92 mg, mp 131-137", and a second crop of 9 mg. The first crop plus 93 mg of AcPhe-Tyr-OMe was dissolved in ethyl acetate and the solution poured through silica gel layered on a sintered glass funnel. Addition of hexane to the initial eluate gave 11 mg of pure Ac-Phe-Tyr-OMe, mp 131-135" (clear melt), which was employed in control 6. Continued elution afforded 120 mg of less pure material, mp 125-134" (unclear melt). Addition of 16 mg of carrier Ac-Phe-Tyr-OMe to the initial 9-mg second crop of Ac-Phe-Tyr-OMe gave a sample that provided Ac-Phe-Tyr after treatment with hydrochloric acid in acetic acid (Mitz et al., 1950). The final sample for control 2 had mp 218-220" after recrystallization from ethyl acetatehexane. The radiochemical and chemical homogeneity of all radiochemicals described was established by the observation that

P E P S I N M E C H A N I S ivl S

each was indistinguishable from its nonradioactive counterpart in hve and tlc and, in the former experiments, a single radioactive peak occurred at the expected position. Technical difficulties encountered in applying these criteria to Ac-PhePhe are discussed later. General Procedure for Transpepfidation Studies. Incubations employed 0.5 M sodium acetate (pH 4.50 i 0.02 and 4.70 =t 0.02)andO.l7~potassiumphosphate(pH1.85 i 0.05) buffers at 35.0 i 0.2” and 0.5 M ammonium acetate buffer (pH 4.70 =t 0.02) at 37.0 f 0.1 ”. For a typical run in an aqueous solvent, 3.3 mg of Ac-Phe-Tyr was dissolved in 1.1 ml of a warm solu= 10 mglml). tion of Ac-m-Phe in pH 4.5 acetate ([Ac-DL-P~~] Dissolution of 2.7 mg of pepsin in 500 pl of the resultant solution, thermostatted at 35 O , initiated reaction. The approximate final concentrations of reagents under “usual” M (assuming conditions were therefore: pepsin = 1.5 X a molecular weight of 34,200); Ac-Phe-Tyr = 8.1 X M; Ac-DL-Phe = 4.8 X M. The simultaneously performed control lacked Ac-Phe-Tyr. Samples of run and control were removed as desired, quenched with an equal volume of ethanol, and frozen pending subsequent hve, tlc, or ninhydrin analysis. The low solubility of Ac-Phe-Tyr-NHz in water necessitated employing a solvent containing 9 methanol for many runs, which were in all other respects identical to those just described. The methanolic incubation mixture for a run was prepared by dissolving 6.2 mg of Ac-Phe-Tyr-NHp in 200 p1 of methanol plus 100 p1 of buffer and then adding 2.0 ml of a solution of Ac-DL-Phe (11.6 mg/ml) in the same buffer. The control lacked Ac-Phe-Tyr-NHz. Determination of the amount of radioactivity incorporated into the acetylated dipeptide relied exclusively on hve separations. Run and control samples were spotted alternately on the hve paper, so each run was adjacent to one or more appropriate controls. At least duplicate analyses of each run were made, where one analysis represented the difference in cpm between one run strip and a sample strip. Duplicate runs were often performed. Typically, Ac-Phe-Tyr-NHz, Ac-Phe-Tyr-OMe, and Ac-Phe-Phe-OMe remained at the original spot, Ac-Phe-Tyr and Ac-Phe-Phe moved to +3.5”, and Ac-DL-Phe was centered at $6.5”. In a good control run, the Ac-DL-Phe peak measured -150,000 cpm while either of the other positions of interest showed 5 3 0 0 cpm (20.273. Papers whose control runs exhibited >600 cpm at the latter were usually discarded but there was an unavoidable tendency toward increased counts per minute at the origin for samples which had been incubated for 2 2 4 hr. Figure 1 illustrates the hve separations achieved. The amount of color produced with ninhydrin reagent in triplicate analyses of run, control, buffer, and solution of Phe,Tyr or Tyr-NH2 in buffer determined the concentration of unreacted peptide in the run sample. This information and the known difference in cpm between run and control permitted calculation of the percentage of unreacted dipeptide that had undergone exchange. The calculation assumed the exchange specific for Ac-L-Phe (see Neumann et a[., 1959, and run 24) and that the sole ninhydrin-positive materials from the substrates were the simple hydrolysis products. The correctness of the latter assumption will now be examined. Ac-Phe-Tyr Ac-m-Phe. Concern about possible complications introduced into the ninhydrin analysis by the formation of Tyr-Tyr (Neumann et al., 1959) led to a semiquantitative tlc investigation of the amounts of Tyr and Tyr-Tyr in the hydrolysis product mixture. This established that: (a) hydrolysis in pH 4.5 acetate under our usual condi-

+

z

l::m:A ,

,

,



,

b ~ 2 - B ’’ ’ ’

;

,

,

;

;

210 ‘239

--D

-

2-

0-

tions produced 7-10x of the theoretical amount of Tyr-Tyr; (b) omission of Ac-DL-Phe from the recipe raised the yield to -3OZ, for pH 4.5 acetate or citrate buffer; (c) 2000 cpm greater for the 6-hr sample. The behavior of this transient peak of radioactivity is indistinguishable from that of synthetic Ac-Phe-Tyr described in control 2, but no experiment performed rules out the presence of Ac-Phe-Tyr-Tyr or higher peptides. ( 5 ) Runs 7-8 established that pepsin from pepsinogen (a

*

194

BIOCHEMISTRY,

VOL.

11,

NO.

2, 1 9 7 2

STODDAKD

gift of Dr. T. P. Stein; see Rajagopalan et al., 1966) behaved similarly to commercial pepsin in our experiment. (6) The failure to detect Ac-Phe-Tyr-NH? from incubations of AcDL-Phe with Ac-Phe-Tyr-NH? might be an artifact introduced because the counting efficiency for paper samples of Ac-PheTyr-NH, is unreasonably low. The satisfactory equilibrations run with Tyr-NH?, control 7, and this control, identical with control 2 except that synthetic Ac-Phe-Tyr-OMe was substituted for Ac-Phe-Tyr, established the unlikelihood of this possibility. At zero time, Ac-Phe-Tyr-OMe had 2017 i 50 cpm and Ac-Phe had 42 i 19. After treatment with pepsin for 24 hr, the figures were 27 =k 13 and 2078 + 91 cpni, respectively. (7) A single check on equilibration experiments was provided by the observation that equilibration of AcDL-Phe, pepsin, and 0.17 M D-Tyr-NHz for 48 hr gave only 66 i 43 cpm more in the run that in the control. The comparable experiment employing L-Tyr-NH? gave >8000 cpm excess in the run, See runs 4-6 of Table 111. Results Tables I and I1 present data on the extent to which synthesis of radioactive substrate oia transpeptidation (eq 1, X’CO = Ac-Phe) has accompanied the simultaneous appearance of hydrolysis products under a variety of experimental conditions. Table I11 describes several determinations of KTII, the equilibrium constant for cleavage of the peptide bond to fully ionized products (eq 2).

Reliabilit), oj the Experimental Datu. The small average errors in the “Acpm” columns illustrate the excellent reproducibility of the hve analytical procedure under optimum circumstances. These errors usually derive from duplicate or triplicate analyses which employed a single electrophoreogram. Runs 21 and 30, the only ones which should definitely have Acpm = 0, define the maximum sensitivity of the hve technique. The small Acpm’s observed in runs 9, 20, 27-29, and 31-33 and equilibration 6 probably reflect a real incorporation of label into substrate for many of them, but Acpm exceeds 100 for none. This ability to achieve Acpm [SI (Denburg et af., 1968; Knowles et af., 1969) and E E Fis~ relatively static when [Ac-Phe] is constant and substrate and hydrolysis products have low and/or similar affinities for pepsin.

+

' 5 1 r 1B '

"

"

IO "

'

"

"

ke + X,X' % S* [XI] kb

Y partitioning ratio, PR, for the intermediate EY between exchange (transpeptidation) and hydrolysis. When applied to Ac-Phe-Tyr and Ac-Phe-Phe, the equation makes the following reasonably true assumptions : (a) hydrolysis is irreversible; (b) any return of EY to S without equilibration of X and X' can be neglected since it is undetectable by the present experiments; (c) [XI] is constant because dilution of Ac-Phe by Ac-Phe from hydrolyzed substrate can be neglected.

Time, Hours

2 : Comparison of the predicted and observed time-dependence for exchange of peptide substrates with Ac-DL-Phe. The curves in A plot [S*]/[So]of eq 6 with the first two sets of values in Table IV: ( 0 ) Ac-Phe-Tyr, (A) Ac-Phe-Phe. Curve B represents 31 X [S*]/ [So] of eq 8 for Ac-Phe-Tyr-NH? if k , = 0.7 hr-1, k./ k, = 1.5 M , and

FIGURE

[x]+ [x']= 3.2 X

M.

B I O C H E M I S T R Y , VOL.

11,

NO.

2, 1 9 7 2

195

SILVER AND

TABLE 11: Extent of the

Runa 21 22d 23. 241 25 26 27 28t 29i 30 311 32 33 34

Pepsin-Catalyzed Exchange of Ac-Phe and Ac-Phe-X in 9

X =

% of So Hydrolyzed

Exptl Detailsb’c

TYr

0

0, 2 7, 3 7, 3 6, 3 6, 4 6, 4 48, 4 0.75, 1 1.25, 2 0, 3 1, 3 1, 6 196 24, 3

Tyr-NHz

44 47 50-60g 51 57 112 20 35 0

42 50

57 106

STODDARD

Methanol at 35 ’.

% of Remaining Acpm -15 zt 32 1778 i 74 1861 rt 119 385 I;t 25 2266 j= 142 2195 rrt 160 65 zt 50 80 46 $1. 39 -8 I;t 9 14 i. 1 22 $1. 20 59 + 15 388 k 67

S Exchanged 22 + 1 25 i 2 27-33h 27 i 2 30 i 2 (loo?) 0.7 0.5 i 0.4 0.9 2 0.1 0.3 i0.3 0 . 8 :k 0 . 3 (loo?)

M in pH 4.5 sodium acetate buffer. c The first number specifies the hr of incubation n As for Table I. b [So] = 7.2-8.0 x and the second, the number of analyses. d Also contained [Ac-~-Phe-o-Tyr-NHd= 8 X IOe3 M. e Also contained [Ac-Phe-OMe] = 1.2 X M. i 2.4 X lo-* h i Ac-L-Phe replaced the usual Ac-DL-Phe. g TIC estimate, since the ninhydrin samples were lost. M. exchange if 50 %hydrolysis, 33 % exchange if 60 %hydrolysis. pH 1.9 phosphate buffer,[EO]= 7.5 X h 27

Solution of eq 3 by a steady-state treatment of [EY] and insertion of the initial conditions [SI = [SO],[Yl = [S*l = 0 provides eq 4-6. First-order plots for the disappearance of unlabeled substrate (eq 4) and total substrate (eq 5 ) determine k, and PR. A single run, when combined with the point for t = 0, is sufficient to define the two parameters. Table IV gives the results of several calculations and Figure 2A compares graphs of eq 6 to experimental data for two examples. The fact that k, for Ac-Phe-Phe is about twice as large as k, for Ac-Phe-Tyr accords well with the report that k,/K,,, for the former is twice that for the latter at pH 2 (Jackson et al., 1966) and strengthens our confidence in this analysis. d[S]/dt

=

-k,[S]

(4)

The approximations of eq 3 are not so good when the hydrolysis of Ac-Phe-Tyr-NHn or Ac-Phe-Phe-OEt is treated. For these cases substrate affinity for pepsin is higher (Denburg et a[., 1968; Knowles et al., 1969) and reversal of products to substrate cannot be ignored (run 34). If equilibration contributes appreciably to [S*], eq 3-6 overestimate PR since they attribute al/ S* to the trapping of intermediate EY. A correction for the contribution from equilibration must be made and eq 7 enables us to do it. Equation 7 approximately

represents the product equilibration phenomenon and possesses the great virtue that the differential equation describing its time dependence is readily integrated if ([XI [X’]) is M) and [S*]